ASSUMING THE WORST MAY NOT BE BAD AT ALL - CARVEDILOL IN HEART-FAILURE TREATMENT

Objective: Carvedilol, a P-adrenoceptor blocking agent with additional al-adrenoceptor blocking properties, has been shown to improve left v entricular function in chronic heart failure (CHF). However, its effec t on mortality has recently been the subject of controversial discussi on. The aim of this meta-analysis is to review the data on mortality f rom two large study programs (the US Carvedilol Heart Failure Study an d the study by the Australia/New Zealand Heart Failure Research Collab orative Group) on additional carvedilol treatment in CHF standard ther apy and to analyse the design and limitations of the individual studie s. Methods and Results: For determination of overall, mortality, all p atients who died and all patients who were withdrawn for other reasons during the open run-in phase of the studies were assigned to the carv edilol group to create a ''worst-case analysis.'' Meta-analysis of mor tality data using the random effects model shows a significantly reduc ed relative risk of 0.55 x 95%-confidence interval 0.325-0.924; p < 0. 05 of death in patients treated with carvedilol compared with patient on standard treatment only. Conclusion: Treatment of CHF using carvedi lol :significantly reduces mortality in patients with CHF, even if the ''worst case'' is assumed by assigning all deaths in the open run-in phase to carvedilol.